On 2014-10-14 17:53:10 -0400, Robert Haas wrote:
> On Tue, Oct 14, 2014 at 4:42 PM, Andres Freund <andres@2ndquadrant.com> wrote:
> >> The code in CHashSearch shows the problem there: you need to STORE the
> >> hazard pointer before you begin to do the LOAD operations to scan the
> >> bucket, and you must finish all of those LOADs before you STORE the
> >> NULL hazard pointer. A read or write barrier won't provide store/load
> >> or load/store ordering.
> >
> > I'm not sure I understand the problem here - but a read + write barrier
> > should suffice? To avoid falling back to two full barriers, we'd need a
> > separate define pg_read_write_barrier(), but that's not a problem. IIUC
> > that should allow us to avoid emitting any actual code on x86.
>
> Well, thinking about x86 specifically, it definitely needs at least
> one mfence, after setting the hazard pointer and before beginning to
> read the bucket chain.
Yes, I can see that now. I do wonder if that's not going to prove quite
expensive... I think I can see some ways around needing that. But I
think that needs some benchmarking first - no need to build a even more
complex solution if not needed.
The solution I'm thinking of is to essentially move away from hazard
pointers and store something like a generation counter per
backend. Which is updated less often, and in combination with other
operations. When a backend need to clean up and sees that there's a
straggler with a old generation it sends that backend a signal to ensure
it sets the latest generation.
Greetings,
Andres Freund
-- Andres Freund http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training &
Services